Equation 9 Pixel Height Mean Difference
2.2 APPLICATION OF REMOTE SENSING AND GIS IN COASTAL STUDIES
2.2.3 SENSOR PLATFORM SELECTION
A multitude of remote sensors have been designed with functions in mind within their resolution competences. The potential of GIS is evident in how we use predesigned instruments and software packages to adapt to study objectives. GIS is flexible and superimposable in any field of study, which only speaks to its interdisciplinary nature. This subsection intends to identify and compare different sensors as well their technological advancements thus formulating the justification for the sensor platforms this study intends to use.
2.2.3.1 OPTICAL SENSORS
As stated, optical remote sensing is dependent on the visible portion of the electromagnetic spectrum, approximately within the 0.4-0.7µm wavelength range (Gibson &Power, 2000).
Images are the result of the combination of the three primary colors red, green, and blue; each resultant color has its own specific wavelength (Ccrs, 2004).
2.2.3.1.1 AERIAL PHOTOGRAPHY
The ability to capture the near-instantaneous state of an environment is the biggest advantage of aerial imagery. They are ideal in scenarios where the spatial resolution takes precedence
over spectral resolution (Ccrs, 2004). What is referred to as IFOV for satellites is called the angular field of view in UAV’s. The orientation of the camera is also of great importance; it can be oblique (side looking) which is ideal for mapping large areas to deduce terrain relief, or vertical which are built for rapid capture whilst controlling geometric distortion (Jeong et al., 2018). Navigation systems are often a part of UAV systems for accurate coordinate capture as well as inbuilt mechanisms that compensate for platform motion. As the UAV moves along the flight line, overlapping images are captured in rapid succession to ensure continuity of the scene for stereoscopic viewing because the overlap of two images means 2 different viewing perspectives, which is ideal in formulating a 3D based stereo model and the ability to infer these geometric measurements is called photogrammetry (Ccrs, 2004).
The biggest benefit of aerial photography is the ability to control the frequency of data collection along with spatial resolution, as well as the easy maintenance and calibration the sensor for optimum spectral resolution (Edwards, 2001). Quality and accuracy are always a concern however, the use of cost-effective digital color and infrared photography has grown particularly in marine based surveys. Digital technologies are more trustworthy for change detection, the use of light sensitive computer chips has done away with the need for digital film and the tedious, costly process of developing them. Images are now easily uploaded to be rectified and analyzed computationally. There is still the need to correct for the conversion of radiance to reflectance for the image mosaicking process as well as geometric correction and verifying ground control points (Malthus &Mumby, 2010).
2.2.3.1.2 OPTICAL SATELLITES
Space borne remote sensing platforms have contributed significantly to advancements in geospatial science. The launch of the Landsat satellite in 1972 made medium-resolution satellite imagery easily available and accessible (Gibson &Power, 2000). Since then, various commercial, high-resolution optical satellites have been launched such as IKONOS and Quick Bird; furthermore, satellites have been launched for the sole purpose of marine observations such as the Coastal Zone Color Scanner in 1978 (Di et al., 2003). Optical based imagery is often used for the extraction of the shoreline in coastal studies (Malthus &Mumby, 2010).
2.2.3.2 LIDAR (Light Detection and Ranging)
LIDAR based sensors operate by transmitting a pulsed laser beam towards a target and its reflectance is recorded (Stockdon et al., 2002). The pulse return time determines the distance between the sensor and the target objects, and this data is captured in the form of a dense point cloud (Ashraf et al., 2011). LIDAR is dependent on factors such as the altitude, attitude, and the scan angle of the sensor to determine the 3D position of each laser beam onto the earth surface (Pe'eri &Long, 2011). The components of a LIDAR system differ slightly depending on whether it is airborne, or ground based. In cases where funding is available, airborne bathymetric LIDAR systems are often the preference for their water penetrating capabilities (Wozencraft &Lillycrop, 2003).
The integral elements of an airborne LIDAR system that make it effective are:
1. Differential-GPS (d-GPS): Calculates the position of the sensor which, also incorporates the scan angle to determine the 3D position of every laser beam on the actual surface
2. Inertial Navigation System: Calculates the attitude of the sensor
LIDAR data has proven to be pivotal in topographic mapping of coastal areas specially to showcase erosion and accretion, which is achieved by subtracting Digital Terrain Models (DTM) (Thiebes et al., 2013). This technology can attain accuracies of up to 5-10cm and several studies have showcased the applicability of LIDAR data in shoreline surveillance (Brock &Purkis, 2009; Deronde et al., 2008; Klemas, 2011; Wozencraft &Lillycrop, 2003).
The greatest advantages of LIDAR data in shoreline mapping are:
1. The ability to adequately map detailed information about the topography, near shore bathymetry and vegetation for a well-rounded perspective on the entirety of coastal environments.
2. The already incorporated statistically formulated tidal datum that allows an output that is truly referenced, and issues associated with using the waterline as proxy are overcome.
LiDAR is the most suitable for shoreline modelling because with a wavelength of 520nm (blue-green laser) it is operational for bathymetric measurements whilst a 1064nm (NIR laser) wavelength is effective for terrestrial topography. LiDAR can thus give a better
depiction of the coastal zone that includes the seabed (Klemas, 2011). This is because electromagnetic energy is greatly reduced with increased wavelength in water.
2.2.3.3 MICROWAVE SENSORS
The 1mm-1m wavelength range of the electromagnetic spectrum affords us a better opportunity to avoid spectral confusion by utilizing radar signals characterized by amplitude and phase (Ccrs, 2004; Filipponi, 2019). The amplitude relates to the strength of the radar response and thus is responsible for reflectance which results in greyscale imagery. The Phase is a single SAR wavelength or a translation of the distance between the sensor and target (Cracknell, 2010). These two factors reconstruct the imaged scene based on signal backscatter or brightness which is determined by the biophysical attributes of the target (Filipponi, 2019). For example, the rougher the surface, the greater the backscatter and the brighter the feature under observation whereas a flat surface would appear dark. The echoed signal can be complex, a multi layered surface would lead to a higher signal proportion (Ferretti et al., 2007).
Whilst this may be visually evident, pixel intensity or digital numbers must be normalised into a computationally detectable format by means of radiometric calibration. Three variations of calibration exist (European Space Agency, 2021; Miranda et al., 2015; Schmidt et al., 2020):
1. Beta (β0) calibration: Reflectivity expressed within slant range (Surface Brightness)
2. Gamma (
γ
0) calibration: Reflectivity associated with the perpendicular plane to the slant range (flattened)3. Sigma (σ0) calibration: Reflectivity expressed within ground range (Surface Backscatter)
Factors such as local incidence angle, dielectric current and polarisation of ground targets also relate to surface reflectivity and conductivity (Kaplan et al., 2021). Dry natural materials have a dielectric constant of 3 to 8 however water has a constant of up to 80. As such moisture in soil or vegetation results in significant reflectivity. The theory of polarisation is based on the rotation of electromagnetic energy as it passes through dynamic materials. This energy is transmitted or received as either horizontal (H) or vertical (V) and helps to measure reflectivity.
Differential Interferometry Synthetic Aperture Radar (DInSAR) is now the preferred approach with radar data in coastal erosion studies. The phase difference between two images taken from two different points of view as a satellite traverses the same orbit, is used to obtain an interference pattern to extract terrain change, deformation patterns and elevation (Ferretti et al., 2007; Mason et al., 1999). The difference between these two views is the baseline. If the baseline is too short, signal phase difference can be undetectable whilst too long of a baseline may introduce additional noise in an image (Prats-Iraola et al., 2015).
DInSAR has been used to map coastal zones as well as generate DEM’s and has the advantages of being able to penetrate shallow waters to capture bathymetric features, as well as being operational in all weather and day or night conditions (Horritt et al., 2001). The temporal lag between the two acquisitions with different angles is still an issue (i.e., low coherence due to long baseline, decorrelation caused by incidence angles impacting the backscattering, and changes in surface roughness from a tidal cycle to another). This limits the use of this method to single-pass interferometry systems for which there is no temporal decorrelation.
Remote sensing is clearly an appropriate tool for coastal monitoring. The variety in technology and data accessibility determines study limitations. The theory associated with beach systems and remotely sensed data is combined below to determine appropriate measurable coastal indicators.